Main

The SARS-CoV-2-specific T-cell epitope repertoire has been studied in some detail1,2,3,4,5,6 (reviewed in Grifoni et al.7); however, comparative in-depth studies of the epitope repertoire targeted by infection- versus vaccine-induced T-cell responses are lacking thus far. Thus, precise prediction of the immune escape potential of emerging VOCs including Omicron/B.1.1.529 from the T-cell response in convalescent individuals compared to vaccinees is hardly possible. In this study, we evaluated SARS-CoV-2-specific T-cell responses in convalescent individuals recovered from SARS-CoV-2 infection (n = 19) as well as individuals after 2 (n = 16) and 3 (n = 7) doses of SARS-CoV-2 vaccination (Pfizer/BioNTech messenger RNA vaccine) (Supplementary Table 1).

Results

SARS-CoV-2-specific T-cell epitope repertoire

We first mapped the overall SARS-CoV-2-specific CD8+ T-cell response against a set of 43 previously described immunodominant SARS-CoV-2-specific CD8+ T-cell epitopes (Supplementary Table 2) restricted by common human leukocyte antigen (HLA) class I alleles1,2,3,4,5,6 in epitope-specific T-cell cultures followed by interferon-γ (IFN-γ) staining. In agreement with their association with a mild COVID-19 course, CD8+ T-cell responses in convalescent individuals targeted most epitopes distributed over all viral proteins, with spike-specific epitopes not being dominant (Fig. 1a, left column). In vaccinees, in contrast and as expected, CD8+ T-cell responses were predominantly directed against spike epitopes (Fig. 1a, right column and Extended Data Fig. 1a). Few CD8+ T-cell responses targeted non-spike epitopes, with the HLA-B*07/N105–113 epitope being the main target. For this epitope, cross-recognition by T cells against common cold coronaviruses has been suggested previously8,9,10. Individual spike-specific CD8+ T-cell epitopes were more often targeted in vaccinees compared to convalescent individuals; the spike-specific CD8+ T-cell repertoire also appeared broader in vaccinees compared to convalescent individuals. When we compared the corresponding viral sequences between wild-type (WT) SARS-CoV-2 variant B and Omicron/B.1.1.529, only a single tested optimal CD8+ T-cell epitope was affected by viral variation in subvariants BA.1 and BA.2 (Fig. 1a, red and Supplementary Table 3).

Fig. 1: CD8+ and CD4+ T-cell responses targeting conserved and mutated epitopes in Omicron/B.1.1.529, BA.1.
figure 1

a, Percentages of CD8+ T-cell responses to previously described optimal CD8+ T-cell epitopes within the complete WT SARS-CoV-2 variant B proteome normalized to the HLA allotype. b, Number and location of spike-specific CD8+ (top) and CD4+ (bottom) T-cell responses to OLPs of the spike protein detectable in SARS-CoV-2 convalescent individuals and vaccinees after two doses of Pfizer/BioNTech mRNA vaccine are depicted. The heatmaps depict the compiled data of tested individuals. Epitopes with amino acid sequence variations in Omicron/B.1.1.529, BA.1 are marked in red. The number of tested individuals (per HLA allotype and in total) and percentage of total T-cell responses targeting variant epitopes are indicated. WT: epitope conserved between WT SARS-CoV-2 variant B and Omicron/B.1.1.529; Mut: epitope mutated in Omicron/B.1.1.529 compared to WT variant B SARS-CoV-2.

Source data

Full size image

Spike-specific T-cell epitope repertoire

To analyse the spike-specific CD8+ T-cell response in convalescent individuals versus vaccinees in more detail, we assessed these responses using overlapping peptides (OLPs) spanning the whole spike protein. For all positive responses, we evaluated the OLP for the described optimal epitopes restricted by the HLA class I alleles expressed by the respective individual. If no matching optimal epitopes were previously described, we performed an in silico analysis to predict the most likely HLA class I restriction and optimal epitope. Using this comprehensive approach, we identified an overall substantially broader repertoire of spike-specific CD8+ T-cell responses in vaccinees (Fig. 1b (second panel) and Extended Data Fig. 1b (second panel)) compared to convalescent individuals (Fig. 1b (first panel) and Extended Data Fig. 1b (first panel)). Indeed, in convalescent individuals, no HLA class I allele restricted more than two spike-specific CD8+ T-cell epitopes, while several HLA class I alleles restricted five or more spike-specific CD8+ T-cell epitopes in vaccinees. In addition, we detected more spike-specific CD8+ T-cell responses per individual in vaccinees compared to convalescent individuals (Extended Data Fig. 1c). Hence, the increased breadth of the spike-specific CD8+ T-cell response in vaccinees was evident at the individual and population levels.

We also analysed the CD4+ T-cell response using spike-spanning OLPs as described above. In contrast to the CD8+ T-cell response, the spike-specific CD4+ T-cell response showed a more limited repertoire of targeted epitopes after vaccination compared to infection (Fig. 1b, bottom). In particular, fewer spike-specific CD4+ T-cell epitopes were restricted by single HLA class II alleles (Fig. 1b, bottom) and fewer CD4+ T-cell responses were detectable per individual (Extended Data Fig. 1d) in vaccinees compared to convalescent individuals. Therefore, the spike-specific CD4+ T-cell repertoire was limited with regard to the individual- and population-based CD4+ T-cell response in vaccinees. Of note, the CD4+ and CD8+ spike-specific T-cell epitope repertoire was relatively stable over time in both vaccinees and convalescent individuals (Extended Data Fig. 2). Importantly, the fewer targeted spike-specific CD4+ T-cell epitopes in vaccinees exhibited high conservation between WT variant B and Omicron/B.1.1.529 SARS-CoV-2 (subvariants BA.1 and BA.2) as it is also the case for most targeted epitopes in convalescent individuals, similar to spike-specific CD8+ T-cell epitopes (Fig. 1b, variant epitopes in BA.1 in red, Extended Data Fig. 1b, variant epitopes in BA.2 in blue and Supplementary Table 3). While the differential responsiveness of CD4+ versus CD8+ T cells to peptide stimulation may limit comparison between the two T-cell subsets, mRNA vaccination appears to particularly broaden and thus increase a CD8+ T-cell response that targets conserved spike epitopes (Extended Data Fig. 1e).

Booster effect on spike-specific T-cell epitope repertoire

Next, to assess the effect of boosting vaccination- or infection-induced T-cell responses by mRNA vaccination on the spike-specific CD8+ T-cell repertoire, we again used overlapping spike peptides to map spike-specific CD8+ and CD4+ T-cell responses in longitudinally followed vaccinees getting their third vaccine dose (Pfizer/BioNTech mRNA vaccine; n = 7; Supplementary Table 1) and convalescent individuals who received an mRNA booster vaccination (n = 3; Supplementary Table 1). After the third mRNA vaccination, we observed a similarly broad and spike cross-recognizing CD8+ T-cell response and similarly limited but still spike cross-recognizing CD4+ T-cell response compared to the completed initial immunization with two vaccine doses (Fig. 2a and Extended Data Fig. 3a). Interestingly, we detected CD4+ and CD8+ T-cell responses targeting more OLPs after the mRNA boost vaccination in convalescent individuals, representing a broader spike-specific T-cell repertoire (Fig. 2b and Extended Data Fig. 3b). Again, the identified CD8+ and CD4+ T-cell responses targeted epitopes that are conserved in Omicron/B.1.1.529. Thus, mRNA booster vaccination increased SARS-CoV-2-specific T-cell responses targeting conserved regions within the spike protein of Omicron/B.1.1.529 in convalescent individuals.

Fig. 2: Boosted vaccine- and infection-induced spike-specific CD8+ and CD4+ T-cell responses.
figure 2

a,b, Number, location and percentages of spike-specific CD8+ and CD4+ T-cell responses to OLPs that are detectable in SARS-CoV-2 vaccinees after the second versus after the third dose of Pfizer/BioNTech mRNA vaccine (bnt162b2, measured 2–4 weeks after vaccination) (a) and in SARS-CoV-2 convalescent individuals who subsequently received a single dose of Pfizer/BioNTech mRNA boost vaccination (bnt162b2, measured 2 weeks after vaccination) (b). The heatmaps depict the data of one representative individual each. Targeted epitopes with sequence variations in Omicron/B.1.1.529, BA.1 are marked in red. c,d, Vaccinees and convalescent individuals with CD8+ (c) and CD4+ (d) T-cell responses within and outside highly conserved selective sweep regions in the spike protein are shown. Statistical analysis was performed with a two-sided Wilcoxon matched-pairs signed-rank test.

Source data

Full size image

Conservation of spike-specific T-cell epitopes

To investigate whether the observed broader spike-specific CD8+ T-cell repertoire after mRNA vaccination may also be beneficial for potentially emerging future SARS-CoV-2 VOCs beyond Omicron/B.1.1.529, we analysed the T-cell response targeting highly conserved selective sweep regions in SARS-CoV-2 that were identified by Kang et al.11 in convalescent individuals versus vaccinees. Selective sweep regions mediate per definition an evolutionary advantage; therefore, it is very likely that newly emerging SARS-CoV-2 VOCs also harbour high conservation within these regions. Four different selective sweep regions have so far been described in the spike protein of SARS-CoV-2 (ref. 11) that also exhibit, as expected, a high degree of amino acid homology among the already evolved SARS-CoV-2 VOCs. For example, complete homology is present in VOC Delta, only one point mutation is present in VOC Alpha or Beta, two mutations in VOC Gamma, seven point mutations in VOC Omicron/B.1.1.529, BA.1 and nine point mutations in VOC Omicron/B.1.1.529, BA.2 (Extended Data Fig. 4a,b). Importantly, compared to convalescent individuals more vaccinees showed spike-specific CD8+ T-cell responses targeting epitopes within the highly conserved selective sweep regions indicating a spike-specific CD8+ T-cell response with focused targeting of highly conserved regions after vaccination (Fig. 2c). A similarly focused spike-specific CD4+ T-cell response was not evident after vaccination (Fig. 2d). Hence, a broadly spike cross-recognizing CD8+ T-cell response is induced after mRNA vaccination that may be also reactive towards emerging SARS-CoV-2 VOCs in the future beyond Omicron/B.1.1.529.

Discussion

In conclusion, our data indicate that (1) convalescent individuals target a variety of SARS-CoV-2-specific CD8+ T-cell epitopes over the complete SARS-CoV-2 proteome with spike-specific CD8+ T-cell responses being non-dominant; (2) in contrast to the CD4+ T-cell response, CD8+ T-cell responses in vaccinees are focused on a broader repertoire of highly conserved spike-specific CD8+ T-cell epitopes leading to an increased cross-recognizing potential; (3) boosting convalescent individuals with mRNA vaccination results in a broader spike-specific CD8+ T-cell response; and (4) CD8+ and CD4+ T-cell responses in both convalescent individuals and vaccinees target epitopes that are highly conserved between WT SARS-CoV-2 variant B, Omicron/B.1.1.529 and potentially future emerging SARS-CoV-2 variants and thus cross-recognize these variants. Hence, our data emphasize the relevance of mRNA vaccine-induced spike-specific CD8+ T-cell responses in combating emerging SARS-CoV-2 VOCs including Omicron/B.1.1.529 and support the benefit of also boosting convalescent individuals with mRNA vaccines.

Methods

Ethics

Patients were recruited at the Freiburg University Medical Center between August 2019 and January 2022. Written informed consent was obtained from all participants. The study was conducted according to federal guidelines and local ethics committee regulations and the Declaration of Helsinki (first revision). The study was approved by the ethics committee of the University of Freiburg (nos. 21-1135 and 21-1372).

Study cohort and clinical definitions

Nineteen convalescent individuals after a mild course of COVID-19 were analysed (Supplementary Table 1). All patients were confirmed to have tested positive for SARS-CoV-2 using PCR with reverse transcription from an upper respiratory tract (nose and throat) swab tested at an accredited laboratory. The degree of COVID-19 severity was identified according to recommendations from the World Health Organization. Moreover, 16 individuals (all testing negative for anti-N-IgGs, (Mikrogen)) were screened 2–4 weeks after the second dose of mRNA vaccination (Pfizer/BioNTech BNT162) and 7 of the same individuals 2–4 weeks after the third mRNA vaccination (Pfizer/BioNTech BNT162). Three additional individuals were analysed who had a mild course of COVID-19 and were vaccinated once with the mRNA vaccine (Pfizer/BioNTech BNT162). The median age of vaccinated donors (n = 16) was 36 years; the median age of donors with a history of natural SARS-CoV-2 infection (n = 19) was also 36 years. The sex ratio of vaccinated donors was M/F 10/6; in donors with a history of natural SARS-CoV-2 infection, it was M/F 11/8. Participants did not receive any compensation for participating in the study.

Peripheral blood mononuclear cell isolation

Peripheral blood mononuclear cells (PBMCs) were isolated from blood samples anticoagulated with density gradient centrifugation (Pancoll Separation Medium; PAN Biotech) and subsequently stored at −80 °C. Frozen PBMCs were thawed in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal calf serum, 1% penicillin/streptomycin and 1.5% HEPES buffer 1 mol l−1 (complete medium; all additives from Thermo Fisher Scientific) until further usage.

Peptides

A total of 180 OLPs spanning the SARS-CoV-2 spike sequence (GenBank accession no. MN908947.3) were synthesized as 18-mer sliding by 7 amino acids and thus overlapping by 11 amino acids with a free amine NH2 terminus and a free acid COOH terminus with standard Fmoc chemistry and a purity of >70% (Genaxxon Bioscience). Since two OLPs contained the two amino acid residues with sequence modification in the ‘stabilized’ mRNA spike vaccine (K986P, V987P), these peptide variants were also synthesized and used, resulting in a total of 182 OLPs. In addition, 60 predescribed SARS-CoV-2-specific optimal CD8+ T-cell epitopes were synthesized. In the figures and Supplementary Table 2, we display only those 43 optimal epitopes that were tested in at least 3 HLA-matched individuals of each cohort (convalescent and vaccinated).

In vitro expansion and intracellular IFN-γ staining using overlapping peptides or predescribed optimal CD8+ T-cell epitopes

In vitro expansion with OLPs or optimal epitopes was performed as follows: 20% of the PBMCs were stimulated with a pool of all 182 SARS-CoV-2 spike OLPs or all 60 optimal epitopes (10 μg ml−1) for 1 h at 37 °C, washed and cocultured with the remaining PBMCs in RPMI medium supplemented with 20 U ml−1 recombinant interleukin-2 (IL-2). On day 10, intracellular IFN-γ staining was performed with pooled OLPs (45 pools with 4 OLPs each). Therefore, cells were restimulated with OLP pools (50 μM), dimethyl sulfoxide as negative control or phorbol 12-myristate 13-acetate and ionomycin as positive control in the presence of brefeldin A and IL-2. After 5 h of incubation at 37 °C, cells were stained for surface markers (CD8+, CD4+; Via-Probe) and intracellular markers (IFN-γ). Subsequently, on days 12–14 single OLPs of positive pools and HLA-matched optimal CD8+ T-cell epitopes were tested by intracellular cytokine staining. Viral amino acid sequences of positive individual OLPs were analysed for predescribed minimal epitopes3,5,6,12,13,14 or the best HLA-matched predicted candidate using the Immune Epitope Database (IEDB, https://www.iedb.org/; we used two prediction algorithms, ANN 4.0 and NetMHCpan EL 4.123, for 8-mer, 9-mer and 10-mer peptides with a half maximal inhibitory concentration of <500 nM). The major histocompatibility complex class I (MHC class I) binding predictions were made using the IEDB analysis resource ANN aka NetMHC v4.0 tool or the IEDB analysis resource NetMHCpan v.4.0 tool. The MHC class II binding predictions were made using the IEDB recommended 2.22 analysis resource consensus tool (smm/nn/sturniolo).

Multiparametric flow cytometry

The following antibodies were used for the flow cytometry analysis: anti-CD8-APC (1:200, SK-1 clone); anti-CD4 eFluor 450 (1:250, RPA-T4 clone); anti-IFN-γ-FITC (1:8, 25723.11 clone); and fixable viability dye (1:200, 1:400, eFluor 506 clone). After cell fixation 2% paraformaldehyde/PBS (Sigma-Aldrich); acquisition was performed on a FACSCanto system (BD Biosciences). Data were collected with the FACSDiva software v.10.6.2 (BD Biosciences) and analysed with FlowJo v.10.0.7r2 (FlowJo LLC). The gating strategy is depicted in Extended Data Fig. 5.

Enzyme-linked immunosorbent assay

Spike-binding antibodies were assessed by anti-SARS-CoV-2 QuantiVac ELISA (IgG) (EUROIMMUN) detecting S1 IgG (<25.6 BAU ml−1: negative; 25.6-35.1 BAU ml−1: marginally positive; ≥35.2 BAU ml−1: positive) according to the manufacturer’s instructions. The SparkControl Magellan software v.2.2 was used for data collection.

Sequence alignment

Sequence homology analyses were performed in Geneious v.11.0.5 (https://www.geneious.com/) using Clustal Omega v.1.2.2 alignment with default settings. The reference genome of human SARS-CoV-2 (MN908947.3; Wuhan-Hu-1 isolate, WT variant B) was downloaded from the National Center for Biotechnology Information database (https://www.ncbi.nlm.nih.gov/). SARS-CoV-2 epitopes were then mapped to the corresponding protein alignment. SARS-CoV-2 VOCs (and subvariants) were identified via CoVariants (https://covariants.org/). Selective sweep regions were marked as described by Kang et al.11. Briefly, Kang et al. analysed a total of 136,114 complete SARS-CoV-2 genomes from the human host. Subsequently to alignment with MAFFT version 7, sweep regions were detected by using OmegaPlus and RAiSD, with the original SARS-CoV-2 isolate Wuhan-Hu-1 genome (NC_045512.2) as an outgroup. Common outliers were manually grouped into 8 regions of at least 50 base pairs; 4 of these were located in open reading frame 1ab (ORF1ab) and 4 in the S region. Selective sweep 1 S323–434, selective sweep 2 S524–545, selective sweep 3 S888–919 and selective sweep 4 S965–1,050 were identified by this method11. The SARS-CoV-2 VOC amino acid sequences were aligned to selective sweep regions and peptides were mapped to the spike protein to identify peptides that localize to the selective sweep regions.

Statistics

Statistical analysis was performed with Prism 9 (GraphPad Software). Statistical significance was assessed by two-tailed Mann–Whitney U-test, two-sided Wilcoxon matched-pairs signed-rank test and Spearman correlation; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.